Research On Scene Modeling Method And Obstacle Avoidance Path Planning Algorithm For Hot-line Working Robot

Hot-line maintenance is essential for the operation and maintenance of smart substations,while also ensuring the power quality of residential and business users.In the past,the maintenance of electrified operations mostly relied on manual labor,the operating environment of transmission lines was complex and dangerous,and the difficulty of operation was high,which had high requirements for the psychological quality and operation ability of personnel.People began to consider robots to replace manual in order to improve the efficiency of electrified operation and reduce the risk index of electrified operation.As a result,a lot of research on hot-line working robots began.In this paper,the primary goal of this research is to find an optimal path without collisions from the start position to the target position in the project of hot-line working robot.At the same time,a method is proposed,which models a transmission line scenario by combining lidar with camera.It provides technical support for the research and development of hot-line working robots.(1)Forward and reverse kinematics analysis of robot manipulator.Firstly,the coordinate system of the robotic arm is established by the modified D-H parameter method,and the D-H parameter table is obtained,which is used to describe the UR10 manipulator.The forward and reverse kinematics formula of UR10 is derived from the homogeneous coordinate transformation matrix.According to the actual working conditions,the joint angle limit is carried out to find the actual reachable space of the arms.The robot is then modeled in MATLAB and the workspace is solved.(2)In view of the shortcomings of the RRT* algorithm,three improvement strategies are proposed to improve the RRT* algorithm.The M-RRT* algorithm(Modified Rapidly-exploring Random Trees Star algorithm)is proposed: The artificial potential field method is used to increase the target bias orientation of random point expansion,so that it expands in the direction of the target point;Taking advantage of adaptive step sizes to improve the efficiency of path planning;Using the quasi-uniform B-spline to smooth the planning path.Using MATLAB to simulate the obstacle avoidance path planning simulation of RRT* algorithm and M-RRT* algorithm in two-dimensional environment and three-dimensional environment,the simulation shows that the improved M-RRT* algorithm takes a greatly reduced time to plan paths,the initial path cost and optimal path cost are lower,and the path obtained is smoother.(3)A scene modeling method for power transmission lines based on the fusion of lidar and camera is proposed.Joint calibration of the relative position relationship between the lidar and the camera,and then preprocessing such as filtering and segmentation of the map point cloud scanned by the lidar to remove unnecessary information.Using Octomap to transform the processed point cloud model into octree model,and only retain its location information.The YOLOv5 s is used to detect the images collected by the camera.Finally,the point cloud data scanned by the laser radar and the camera image data are fused in coordinates,facilitating the robot to perceive the surrounding environment and identify obstacles for obstacle avoidance path planning.(4)An experimental scenario for obstacle avoidance path planning for Hot-line working robots was built,using Move It! Configure a motion planning group for the robot,communicate the real robot with the robot ontology model under the ROS environment in the upper computer,and send the path information planned by RRT*and M-RRT* to the real robot.Controlling the motion of the robot arm to the target point in simple and complex environments,and recording the changes in the angle of six joints and the position and posture of the end during the motion of the robot arm to verify the smoothness of the robot arm’s trajectory.Comparing the Cinit,Tinit,Cbest,and Tbest of the RRT*,Informed-RRT* and M-RRT* algorithms,and take the average of 10 experiments to verify the effectiveness and feasibility of M-RRT* algorithm performance improvement.The experimental results show that compared to RRT*algorithm,the improved M-RRT* algorithm improves the time retrieval efficiency of the initial path and the optimal path by 88.4% and 85%,respectively,and increases the initial path cost by 11.4% and the optimal path cost by 0.88%.Compared with the Informed-RRT* algorithm,the initial path acquisition time and optimal path acquisition time of the M-RRT* algorithm are reduced by 86.1% and 79.4%,and the initial path cost and optimal path cost are reduced by 6.7% and 0.63%,respectively.